During a critical period of early postnatal development, an asymmetry in the quality of visual input to the two eyes shifts ocular preference away from the weaker eye and induces amblyopia, the most common cause of monocular visual deficits in humans. Amblyopia is highly resistant to reversal in adulthood, due in large part to th termination of the critical period of heightened plasticity. Understanding how the enhanced plasticity of the critical period is initiated and terminated over development is fundamental to th development of therapeutic strategies aimed to reactivate plasticity to treat amblyopia in adults, which can be translated to a clinical population and to other critical periods. A popular model for the regulation of the critical period proposes that inhibitory control of plasticity at excitatory synapses is mediated by the maturation of the output of fast-spiking interneurons (FS-INs) that mediate perisomatic inhibition. However, we have shown that ocular dominance plasticity can be induced several months after the maturation of perisomatic inhibition. We propose instead that ocular dominance plasticity is regulated by plasticity upstream of inhibitory output, likely affecting the recruitment of inhibition into functional circuits. In addition we propose that the functional connectivity of Pyr->FS synapses must be retained in a permissive range for ocular dominance plasticity to be expressed, as larger reductions in Pyr->FS connectivity induced by genetic manipulations inhibit the expression of ocular dominance plasticity. Our preliminary analysis of the regulation of excitation from pyramidal neurons onto FS-INs (Pyr->FS) reveals that monocular deprivation during the critical period may functionally disconnect FS-INs from the cortical network by significantly reducing the number of excitatory inputs onto these neurons. Therefore we hypothesize that a novel mechanism of plasticity, deprivation-induced loss of functional Pyr->FS connectivity 1) is an early and obligatory step in the shift in ocular dominance induced by MD, and 2) determines the timing of the critical period. We propose a multidisciplinary set of experiments to test these hypotheses that combine: the expertise of the Quinlan lab in the examination of physiological changes in visual cortex in vivo in response to monocular deprivation; the expertise of the Kirkwood lab in the direct assessment of contribution of changes in single synapses to activity-dependent plasticity in the visual cortex; and the expertise of the Lee lab in the use optogenetic methods to identify foci and mechanisms of activity-dependent changes in synaptic function. Our model for the regulation of the timing of the critical period refutes many widely-held assumptions regarding developmental changes in synaptic plasticity in the mammalian cortex.